What is truth?

[flipzagging]

Well, we've seen a few waves of important news breaking on social media just in the last week. Most notably the Iran election, but there was also the Honduras coup. Both were stories that happened to a lot of people at once, where even the central narrative was highly disputed, and the 'official' sources of information were all interested parties

Comments

[leolo]

You have a way of positing unsolvable problems. I'm reminded of your post a few years back about how to make a "perfect" web dating site.

I'll throw in a few data points : on Penn and Teller's Bullshit television series, they had a show about Jesus (I think). They used the example of Elvis' fav food : fried chicken. Elvis was dead 25 years (at the time) but we don't know the "real" recipe for Elvis' fried chicken. They also repeatedly point out that many believe "Elvis didn't do no drugs."

In other words, there is a very small window during which something may be "known". We can both look at a thermometer and agree "yep, it reads 25C." But next week/next decade we might no longer remember. Most records can be falsified (cf Bush's military service) With enough thought, we can delude ourselves about anything (cf the first-hand witnesses to the faked Moon landing "cover-up").

So basically, we're doomed to never knowing for sure.

[flipzagging]

So are you saying that nobody knows anything for sure? That's kind of a boring objection.

I'm not asking for a more perfect standard of truth. I was asking (generally) if the level of verification that say, a well-paid and skeptical editorial staff does today, could ever be duplicated using peer production. It seems to me that it could be improved on, but let's just say equalled for now. We're just talking about a task that used to be done one way being done in another way. I see no theoretical impossibilities there.

[leolo]

First, I'm saying that the truth about an event is harder to know the further (distance, time) we get from the event.

Second, your reformulated problem: imma chew over it a bit.

[fg]

the kind of framework required to assemble higher truths from fuzzy atomic truths, and some kind of authority system, there be dragons in that math. but it's a very compelling idea :)

a big part of the problem is the feedback loop. you need to bring some outside information - a kudos system alone promotes islands of truth where people are promoted because they encourage agreement, not because the facts they produce are true. collective dissonance :)

which is kind of what we have in place now, in a non-formal way. a wobbly, tenuous trust system that is massively non-homogenous. if you want a tribe that supports your theories about the faked moon landing, you'll find them.

even more devilish, did you see the guardian article about the social hack where a university student inserted fake news into major media outlets? see the guardian posti waffle as to whether or not a tool like this would be helpful. i imagine you cited a political example because you believe that if people knew the truth, they'd react more strongly

[fg]

but having said that, if this is your greenfield project, count me in :)

[flipzagging]

I wasn't even thinking as far as motivating social change. I was really only thinking about truth as a good in itself and how do we achieve that. There are also a lot of cases lately where the public is asked to swallow complex, new doctrines to solve various problems (health care, terrorism, global finance) and largely, we don't have anybody evaluating the truth of these schemes, only their political viability

[fg]

jah, the stuff i am doing now is kind of related. i will shoot you an email later.